Redis

What is Redis?

Redis is an in-memory data store. It keeps its dataset in RAM rather than on disk, which is what makes it so fast — typical operations complete in under a millisecond. It's used as a cache, a session store, a message broker, a job queue, and sometimes as a primary database for the right kinds of data. If you need something faster than a relational database can provide, Redis is usually the answer.

Data structures

Redis isn't just a key-value store. It supports strings, hashes, lists, sets, sorted sets, bitmaps, and more. This means you can use native Redis operations to do things that would require complex SQL queries elsewhere — ranked leaderboards with sorted sets, unique visitor counts with HyperLogLog, pub/sub messaging with the native pub/sub system. The data structure you pick determines what operations you can run and how efficiently.

Caching

The most common use of Redis in web applications is caching. You store expensive query results, rendered HTML fragments, or API responses in Redis with an expiry time, then serve them from Redis instead of recomputing them on every request. The speed difference between Redis and a database query is typically an order of magnitude or more, which makes a meaningful difference under load.

Queues and pub/sub

Redis is widely used as a job queue backend. Tools like Sidekiq (Ruby), Laravel Horizon (PHP), and Bull (Node.js) use Redis lists or streams to store jobs that workers pick up and process asynchronously. The pub/sub system lets services publish messages to channels and have other services consume them in real time, which is a lightweight alternative to a full message broker for simpler use cases.

Persistence and deployment

Despite being in-memory, Redis can persist data to disk using snapshots or an append-only log. This means it can survive restarts without losing data, though the recovery time depends on the size of the dataset. Redis Sentinel and Redis Cluster provide high availability and horizontal scaling for production deployments.